When should you stop a poorly selling marketplace product?

A product that sells little is not automatically a bad product. It may barely appear for relevant searches. Customers may see it, but find the price, delivery promise, or main image unconvincing. Sometimes the conclusion is simpler: the product does not solve the problem well enough or does not fit the marketplace.

The mistake is treating all these situations with more advertising budget.

Find where the chain breaks

I assess a product in this order:

1. Relevant impressions. Does the product appear for searches and categories where its intended customer actually buys? Irrelevant impressions are noise, not reach.
2. Click-through rate. With relevant visibility, CTR indicates whether price, imagery, reviews, delivery, and proposition are attractive at first glance.
3. Conversion. If people click but do not buy, the issue usually sits deeper in the listing, comparison, price-value relationship, or product itself.
4. Contribution margin. What remains per order after cost of goods, marketplace fees, fulfilment, returns, and advertising?
5. Scalability. Can additional volume demonstrably improve purchasing, logistics, reviews, and organic position, or does it simply increase the loss?

This order prevents a weak listing from being mistaken for a weak product, while also preventing months of spend on a product with little relevant demand.

Four possible decisions

Invest

Relevant demand exists, the first click response is healthy, and unit economics can work at sufficient volume. A focused investment in content, price, stock, and advertising may be justified.

Run a controlled test

There are positive signals but not enough evidence. Do not change everything at once. Test price first, for example, followed by the main image or another keyword segment, with a predefined budget and end date.

Maintain with minimal input

The product covers its costs and needs little attention, but scaling is not attractive yet. Keep it available, automate the operation, and direct your time toward products with stronger potential.

Stop

After a fair test period, relevant CTR, conversion, and contribution margin remain weak, while listing quality, delivery, and price position have been tested properly. Stopping is not failure. It releases capital and attention for a better opportunity.

Why three to six months is often enough

For many marketplaces, three to six months provides enough evidence to choose a direction, provided the product actually received visibility and the fundamentals were sound. This is not a universal deadline. Seasonal products, long purchase cycles, and emerging marketplaces may require more time.

I have seen marketplaces that simply did not become viable for a brand. Fully exiting is one option. Another is a break-even presence with minimal input, because an algorithm or assortment change can make the channel relevant later.

Advertising budget is not the scarcest resource

Time, attention, and working capital usually are. The right question is not only: could this product ever sell? It is: will the next hour or euro produce more here than in the best available alternative?

That is why a Marketplace Opportunity Scan should not only identify opportunities. It should also show where further investment is not justified.

Sources and context

The exact algorithmic weights used by bol.com and Amazon are not public. Bol does state that price, delivery conditions, availability, and quality influence its buy box. Amazon likewise treats selling fees, fulfilment, and advertising as combined inputs to profitability in its Revenue Calculator guidance.